samyukta.mogadali

 Profile Answers by samyukta.mogadali 

• Jul 28th, 2010

 

Low Confidence: Normally when stats are difficult to use precisely

High Confidence: Normally Optimizer is sure of the results based on the stats available

yuvaevergreen

 Profile Answers by yuvaevergreen 

• Aug 13th, 2010

 

The explain generally displays the plan which would be prepared by Optimizer.
We can say that(though not accurate), plan would be the steps followed by Teradata to solve a query.
Confidence levels indicate how well the optimizer is aware about the  demographic data for a particular step.
High confidence: Means the optimizer know about the no of rows that would be returned as a result of that step. Examples would be, when  PI statistics exist, when the column or range stats exist or no join involved.
Low confidence: Some stats available. Join and stats available on both sides of the join.
No confidence: No stats available. Join involved.